Effect Moderation and Mediation in Prevention Programs While school and community based tobacco and drug abuse prevention programs based on social influences have produced the greatest cumulative evidence for prevention effectiveness, their effects are inconsistent. One possible cause of this inconsistency is effect moderation such that factors either intrinsic or extrinsic to some young people make them refractory to the interventions. In the first 4 years of the USC TTURC we carried out 2 longitudinal trials testing social influences prevention programs in a wide variety of cultural contexts in southern California and central China.We found evidence in both countries for a dispositional attribute moderator effect such that youth scoring high on hostility or depression experienced reductions in smoking measured 1 and 2 years post program, but those scoring low did not. In California we also found a moderator effect for social context, with a progam containing collectivist content (act for the good of others) effectively reducing smoking in homogeneous Hispanic schools but not in multi-cultural schools, and a program with an individualist approach (look after yourself) vJorking in multi-cultural schools but not homogeneous Hispanic schools. We propose a model whereby dispositional phenotypes interact with social context (programs) to alter response to prevention programming, and do so by working through mediation pathways involving cognitive, affective, and motivational arousal. Likewise we propose environment X environment interactions such that features of environment render one more or less receptive to a prevention program. Socio-cultural and dispositionai variations in populations might account for many of the inconsistencies in prevention program effects. We propose 3 studies: 1) a quantitative analysis of effect mediation by dispositional attributes and effect mediation by cultural context in 4 trials previously carried out in diverse cultural, regional, and international settings, 2) rigorous development of better phenotype an socio-cultural measures to better assess effect moderation and mediation. 3) a pilot trial designed to replicate the dispositional moderation effect and, through high intensity interventions negate, the effect by reaching otherwise refractory youth. Project 3 will do program by genotype assessments on one of the previous trials for comparison with program by phenotype analyses in this project.